University of Pittsburgh Emergency Medicine

Department of Emergency Medicine the University of Pittsburgh

University of Pittsburgh Emergency Medicine

www.emergencymedicine.pitt.edu


Chair, Department of Emergency Medicine;

Professor of Emergency Medicine, Medicine, and

Clinical and Translational Sciences,

School of Medicine, University of Pittsburgh

Donald M. Yealy, MD

Early Septic Shock Care –Phenotypes and How We Got Here


Financial disclosures

Deputy Editor, Annals of Emergency Medicine

My external funding past 36 monthsNHLBI – PETAL Network (PI); T32 Research Training GrantNIGMS - RO1 ProACT (procalcitonin in LRTI)Royalties from three texts (Tintinalli’s Study Guide; The Trauma Manual; ED Critical Care) and UpToDate (not sepsis related)

Expert opinions - civil



Learning Objectives

List three reasons for the ProCESS and other two non-US trials

Note two main outcomes

List two similarities in design and outcome with the more recent trials

State how different sepsis phenotypes exist and could influence actions and assessments.


Origins of sepsis

Hippocrates (BC ~460-370)Sepsis (σῆψις)

The process by which flesh rots, swamps generate foul airs at night, and wounds festerIt is rank, disease-producing, and evil


The 2000 year evolution of ‘germ theory’

Fracastoro (1478-1553)Passage of minute bodies from one person to another

Pasteur (1822-1895)Confirmation of the germ theory; vaccination

Semmelweis (1818-1865) and Lister (1827-1912)Antiseptic control

KochScientific basis for interrogation of mechanism of action

Domagk, Fleming, et al (20th century)Modern era of antibiotics


Bone et al. Ann Intern Med 1992; Levy et al. CCM 2003

What do we think sepsis ‘is’?

Patients still die DESPITE effective antibiotics

Sepsis is a host response to infection gone awry!A case of harm by friendly fire

When organs fail, the sepsis is called ‘severe’1992 and 2003 International Consensus Definition

TraumaBurns

Virus

Fungus

Bacteria

Parasite

Sepsis

Sepsis

Hostinflammatoryresponse

Infection


Sepsis accounted for 36-55% of all hospital deaths

#1 cause of hospital deaths in the US

9


Landmark – 11k (!!) citations on Google Scholar

Single center studyN=263

Protocolized EGDT vs. usual care after early detection/fluid bolus

16% absolute mortality reduction30% vs. 46%Exp arm: 5L (vs 3.5 L IVF), more blood, dobutamine

Background


Protocol-based EGDT

Early insertion of CVC for ScvO2

Titrate to CVP, MAP and ScvO2

Potential PRBC and dobutamine

use


Residual questions

Is the difference due to the act of ‘protocolizing’ or attention only to resuscitative care after early/better identification of shock?

Are all elements of the protocol necessary?Early central venous catheterization in all patientsCVP guided initial fluid therapyScvO2 monitoring to guide therapy, notably red cell transfusion and dobutamine

Are the results generalizable?Now?In broader multicenter settingFollow-up EGDT studies often used “off/on” design with limited CVC/protocol adherence – testing attention instead of protocol?


The ProCESS trial

Early septic shock in the Emergency Department

Randomization to one of 3 arms for 6 hours of resuscitationProtocol-based ‘Early Goal-directed Therapy’Protocol-based standard therapyUsual care

Two sequential hypotheses …Is protocol-based resuscitation superior to usual care?If so, is EGDT superior to protocol-based standard therapy?

Where ‘standard’ therapy does NOT includeCentral venous pressure and oxygen monitoringLatter guiding blood or dobutamine



Resuscitation from randomization to 6h

Intravenous fluidsEGDT protocol

2.8 L p<0.001

PST 3.3 LUsual care 2.3 L

Intravenous antibiotics

EGDT protocol

97.5% p=0.90

PST 97.1%Usual care 96.9%

Intravenous fluids Vasoactive agents and blood

Pressors Blood Dobutamine


Outcomes

No difference in mortalityResults unchanged when adjusting for potential site heterogeneityHigher dialysis-dependent renal failure in protocol-based Std Rx arm


Mortality over time

EGDT protocol

PST protocol

Usual care

In-hospital mortality up to 60 days

Mortality up to one year


Secondary outcomesEGDT PSC Usual

Higher ICU use with EGDTPossibly due to monitoring differences


A priori subgroup analyses

No interaction between treatment arm and …AgeSexRaceSource of infectionType of shock

True for 60d hospital mortality, 90d mortality, and 1ymortality

Process of care hour +6-72 – no difference (“catch up” or correcting gaps not evident)




Put all 3 Together…..


And put together again…..


Conclusions

For patients presenting with early septic shock in the setting of …

Prompt recognitionPrompt intravenous fluid bolus for hypotensionPrompt intravenous antibiotics

… there is no superiority to routine…Protocol-based resuscitation if other aggressive recognition/care existsMandatory central line placement in all patientsScvO2 monitoring, with triggers for blood transfusion and dobutamine


Impact

Looking early and hard, treating aggressively with antibiotics and hemodynamic support, and rechecking is more important than “how” support is done

“EGDT” vs ‘egdt’ (one set of specific goals vs. concepts)

The durable message from Rivers through ProCESS – septic shock is deadly, and early care matters.

Not an anomaly though still opportunity - outcomes matchother reports (Jones JAMA 2010; Kakonen JAMA 2014)


So, buckets of volume then pressor, right?

Maitland (NEJM 2010) – those with more fluids did worse than rest

Andrews (JAMA 2017) – same storyPatients (children/acute and chronic illnesses)Ancillary care (esp. ICU, ventilation)PathogensAntimicrobial therapy

CLOVERS – NHLBI trial through PETAL network – RCT of early fluids vs early pressor in US sepsis care. All get @ 2L IV, then restrict/pressor or more fluids.


Antibiotics

All US/European/Australian trials gave early (inside hours)

How early is early enough?

Before resuscitation? Lab data conflict

Best human data – Seymour et al NEJM, NY state observational cohort

Earlier matters – but crude measure (time to complete 3 hrbundle – better if done < 12 hrs)Each hour delay increases mortalityLater giving sites do worse


Most Recent Data – Seymour et al


Results

The mandated efforts improved outcomes – mortality and others

Outcomes improved in non-mandated settings also, but not as much relatively

Which parts are less clear – recognition and antibiotics still key, latter = sooner

Who benefits the most from what? Phenotypes…


Newest Data – Seymour et al


Findings

4 derived/validated phenotypes in > 60k episodes:the α phenotype was the most common (n = 6625; 33%) and included patients with the lowest administration of a vasopressor;

in the β phenotype (n = 5512; 27%), patients were older and had more chronic illness and renal dysfunction;

in the γ phenotype (n = 5385; 27%), patients had more inflammation and pulmonary dysfunction;

and in the δ phenotype (n = 2667; 13%), patients had more liver dysfunction and septic shock


So what?

In the derivation cohort, cumulative 28-day mortality was:287 deaths of 5691 unique patients (5%) for the α phenotype;

561 of 4420 (13%) for the β phenotype;

1031 of 4318 (24%) for the γ phenotype;

and 897 of 2223 (40%) for the δ phenotype.

Across all cohorts and trials, 28-day and 365-day mortality were highest among the δ phenotype vs the other 3 phenotypes (P < .001).


Sepsis Phenotypes – what next?

Are there genotypic or other patient based observations –at any level – that tie to these phenotypes? Can they be measured in a useful way?

Are phenotypes stable, and when can one be assigned?

How do the common interventions interact with outcome across phenotypes?

What should we do when assessing quality or mandating care with phenotypic information?


SEP-3 : Third International Conference


SEP–3 and qSOFA


SEP-3 and EM

No participation – acute, early view under-represented

Benefits vs harmSimpler gradation – no “severe sepsis” any moreSeptic shock now only with elevated lactate and vasopressor use (restrictive)? More reproducible

Sensitivity vs specificity, validationqSOFA robust In ED – unknown performance vs “old way” (SIRS plus old definitions, or gestalt) – likely limited sensitivity in early stagesOverall effect (change # with “sepsis” and “septic shock”, and deaths attributed – but will fewer die with infections?)


The “New one-hour bundle” from SSC

Released 2018

Three from SSC group created this new bundle (even more selective)

Targets one hour actions – notably volume + antibioticsTime 0 = arrivalStarting vs finishing“Aspirational” and “we know the starting time is wrong”

Impact?OveruseUse in non-sepsisIs all sepsis the same?


Our take away

Sepsis kills, still

Key is looking early and often – not “one test/thing”.The field and the ED matters

Use tools to aid – order sets, complimentary tests

Trying matters – ATB and restoring perfusion (LR or pressors) more than specific steps – assess and re-assess

Sepsis and care has many faces – soon, we will tailor care better – even the simple things like volume, antibiotics and pressors.



Questions ?

Documents

University of Pittsburgh Emergency Medicine